Molecular dynamics simulation of vibrational energy relaxation of highly excited molecules in fluids. I. General considerations

Abstract

Methods of implementation of classical molecular dynamics simulations of moderate size molecule vibrational energy relaxation and analysis of their results are proposed. Two different approaches are considered. The first is concerned with modeling a real nonequilibrium cooling process for the excited molecule in a solvent initially at equilibrium. In addition to the solute total, kinetic, and potential energy evolution, that define the character of the process and the rate constant or relaxation time, a great deal of important information is provided by a normal mode specific analysis of the process. Expressions for the decay of the normal mode energies, the work done by particular modes, and the vibration–rotation interaction are presented. The second approach is based on a simulation of a solute–solvent system under equilibrium conditions. In the framework of linear nonequilibrium statistical thermodynamics and normal mode representation of the solute several expressions for the rate constant are derived. In initial form, they are represented by integrals of the time correlation functions of the capacities of the solute–solvent interaction atomic or normal mode forces and include the solute heat capacity. After some approximations, which are adequate for specific cases, these expressions are transformed to combinations of those for individual oscillators with force–force time correlation functions. As an attempt to consider a strongly nonequilibrium situation we consider a two-temperature model and discuss the reason why the rate constant can be independent on the solute energy or temperature. Expressions for investigation of the energy redistribution in the solvent are derived in two forms. One of them is given in the usual form of a heat transfer equation with the source term describing the energy flux from the excited solute. The other form describes the energy redistribution in the solvent in terms of capacity time correlation functions and can be more convenient if memory effects and spatial dispersion play an important role in energy redistribution in the solvent.

Watching Photoinduced Chemistry and Molecular Energy Flow in Solution in Real Time

“…[46,106,107,111,112,121] In general, x(w) has a maximum at very low frequencies and decays exponentially with increasing w, such that VET is dominated by the lowest-frequency modes of the solute. [111,122] However, it is also well-known that collisions that are characterized by a low order are favored, that is, only states of the molecule that differ by a small number of quanta and that excite the least number of quanta in the solvent are accessible. It is thus the trade off between a low-order process and a small energy gap that is important for effective VET.…”

“…[46,112] Beyond such an approach, flexible MD calculations should be able to treat IVR and VET in solution, even for polyatomic systems. Recently, VET of azulene in solution and supercritical fluids [122][123][124] and VET as well as IVR of diiodomethane in solution was investigated in full MD simulations. [96] The latter study has accompanied our recent experimental investigations of the intra-and intermolecular vibrational energy flow of CH 2 I 2 in solution.…”

Echtzeit‐Beobachtung photoinduzierter Chemie und molekularen Energietransfers in Lösung

“…106, 107, 111, 112, 121 Im Allgemeinen weist ξ ( ω ) ein Maximum bei niedrigen Frequenzen auf und nimmt mit zunehmendem ω exponentiell ab, sodass die niederfrequenteste Mode des Solvats den VET dominiert 111. 122 Zugleich werden jedoch solche Übergänge bevorzugt ablaufen, die einerseits mit einer geringen Änderung der Quantenzahlen einhergehen und bei denen andererseits nur wenige Schwingungsquanten des Lösungsmittels angeregt werden – also Übergänge niedriger Ordnung. Dieses Wechselspiel zwischen der Ordnung der Übergänge und der zu überwindenden Energielücke ist ausschlaggebend für einen effektiven VET.…”

“…112 Darüber hinaus sollten MD‐Rechnungen jedoch auch für IVR‐ und VET‐Prozesse mehratomiger Moleküle in Lösung anwendbar sein. Kürzlich wurden vollständige MD‐Simulationen des VET von Azulen in überkritischen Fluiden und in Lösung122–124 sowie zu VET und IVR von Diiodmethan in Lösung beschrieben 96. Letztgenannte Studie begleitete unsere Experimente zum intra‐ und intermolekularen Schwingungsenergietransfer von CH 2 I 2 in Lösung.…”

Molecular dynamics simulation of vibrational energy relaxation of highly excited molecules in fluids. III. Equilibrium simulations of vibrational energy relaxation of azulene in carbon dioxide

“…Over the last years vibrational energy relaxation ͑VER͒ in liquids has been studied for a variety of systems both experimentally and theoretically [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] leading to an understanding of many factors affecting this type of process like oscillator frequency, solute-solvent interaction, mode of motions, temperature, and the significance of quantum mechanical aspects. But while most theoretical investigations have been dealing with small molecules consisting of only a few atoms, data are available [21][22][23][24][25] now for moderate size molecules, too.…”